A near-tight lower bound on the density of forward sampling schemes
2024
A Lower Bound on the Density of Forward Sampling Schemes
publication
Evidence: high
Author Information
Author(s): Kille Bryce, Groot Koerkamp Ragnar, McAdams Drake, Liu Alan, Treangen Todd J
Primary Institution: Rice University
Hypothesis
How close are existing sampling schemes to achieving minimum density?
Conclusion
The study proves a near-tight lower bound on the density of forward sampling schemes, indicating that existing schemes are much closer to optimal density than previously known.
Supporting Evidence
- The new lower bound closes most of the gap between existing schemes and optimal density.
- With current default settings, the best known scheme is at most 3% denser than optimal.
- The study provides a framework for future research to improve sampling scheme metrics.
Takeaway
This study shows that when sampling sequences, we can do it more efficiently than before, and existing methods are already very close to the best possible way to do it.
Methodology
The authors developed a new lower bound on the density of forward sampling schemes and compared it with existing methods using integer linear programming.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website